ATi tries to redefine MAXXimum G-Force

Yesterday both Sharky Extreme and FiringSquad threw up previews of the ATi Rage Fury MAXX. The MAXX is essentially two Rage 128 Pro chips on the same board, combined with 64(!) megs of memory. Preliminary results show it more or less neck and neck with nVidia’s GeForce 256 in some cases, and just about wiping the floor with it in others. One example from Sharky’s page: in Q3A 32-bit, the MAXX squeaks past the GeForce at 800×600 with 71.4 fps to GeForce’s 68.5. Bump the resolution up to 1600×1200, however, and the MAXX whomps the GeForce 23.3 fps to 16.2. But I digress. Go check out the previews for yourself. Here’s the Sharky Extreme preview and the FiringSquad preview.

Comments closed
    • Anonymous
    • 17 years ago

    What is the price of a gforce fx????

    • Anonymous
    • 20 years ago

    *[Anon@bowerbird.cc.uq.edu.au<]* If you look at the tests the MAXX maybe whooping along its doing easy thing (like Quake II) but when there is thinking involved and the GPU kicks in from the GeForce, and the MAXX doesnt hold up as well. b[

    • Anonymous
    • 20 years ago

    *[

    • Anonymous
    • 20 years ago

    *[Anon@sdn-ar-001gasavaP205.dialsprint.net<]* Multichip parallelism is something already in widespread use today in supercomputers. The only reason we haven\'t seen any more of it than we have in the general x86 desktop market is because technologies are only now just beginning to reach the point where singular chip design starts running into limits on size, not to mention economics. When graphics chip makers go to .18 Microns, they are getting very close to the maximum minituarization limits inherent in silicon technology. Intel\'s chief scientist recently pointed this out as a not-to-distant problem for x86 cpus. If it\'s a problem for a company as well-heeled as Intel, it\'s doubly, triply a problem for the much smaller graphics chip makers. This is something you can expect to see a lot more of in the future, you can bet on it. Besides, what difference could it possibly make if they use one, two, or three chips? If the designs are good and the latencies are low, they can do incredible things without running their chips at 600MHz (buring most of them up in the process), or demanding extremely expensive, exotic ram configurations. What this means for the consumer is a big improvement in performance at a low price. Why pay 2X for a card, when you can buy a card for X that outperforms it by 20-50%? Isn\'t that what it\'s all about anyway?

    • Anonymous
    • 20 years ago

    *[Anon@131.107.3.84<]* Great idea, Too bad they had to stick 2 third rate chips together to merely match the current high end. If a company with better 3d chips like matrox or nvidia had done this it would be far more impressive. I wonder if ATI\'s method will work with the coming embedded framebuffer dram chips?

Pin It on Pinterest

Share This